HIV, HBV, HCV and syphilis blood testing using BIO-FLASH technology. based algorithm before gastrointestinal endoscopy

Transcription

1 JCM Accepted Manuscript Posted Online 5 October 2016 J. Clin. Microbiol. doi: /jcm Copyright 2016 Jun et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license. 1 2 HIV, HBV, HCV and syphilis blood testing using BIO-FLASH technology based algorithm before gastrointestinal endoscopy Authors: Zhou Jun, a # Chen Zhen, a Zhang QuiuLi, a An YuanQi, a Verónica Vocero Casado, b Yuan Fan, a Affiliations: a Department of Blood Transfusion, Beijing Military General Hospital of PLA, Beijing, China; b Marketing Department. Biokit. Lliçà d Amunt. Barcelona, Spain Corresponding author: Zhou Jun, Department of Blood Transfusion, Beijing Military General Hospital of PLA, Beijing, China. Tel.: ; zhoujunlz@163.com Author information: Chen Zhen: @163.com; Tel: Zhang QiuLi: bjssjszql2008@126.com: Tel: An YuanQi: @139.com; Tel: Yuan Fan: @qq.com; Tel:

2 21 Abstract Currently, conventional enzyme immunoassays which use manual gold, and colloidal tests (GICT) are used as a screening tool to detect Treponema pallidum (syphilis), hepatitis B virus (HBV), hepatitis C virus (HCV) and human immunodeficiency viruses (HIV 1/2) in patients undergoing surgery. The present observational, cross-sectional study compared the sensitivity, specificity and workflow characteristics of the conventional algorithm using manual GICT with a newly proposed algorithm that uses automated BIO-FLASH technology as a screening tool in patients undergoing gastrointestinal (GI) endoscopy. A total of 956 patients were examined for the presence of serological infection markers (HIV-1/2, HCV, HBV, and syphilis). The proposed algorithm using BIO-FLASH technology was superior in the detection of all markers (100.0% sensitivity and specificity for anti-hiv and anti-hcv antibodies, HBV surface antigen [HBsAg] and syphilis detection) compared with the conventional algorithm based on the manual method (80.0% sensitivity and 98.6% specificity for anti-hiv; 75.0% sensitivity for anti-hcv and 94.7% for HBsAg, and 100% specificity for both markers) in these patients. The automated BIO-FLASH screening algorithm also reduced the operation time by 85.0% (205 minutes) per day, saving up to 24 hours/week. In conclusion, the use of the newly proposed screening algorithm based on the automated BIO-FLASH technology can provide an advantage over the conventional algorithms based on manual methods in the screening of HIV, HBV, HCV and syphilis before GI endoscopy. Keywords: automated BIO-FLASH technology, gastrointestinal endoscopy, hepatitis B virus, hepatitis C virus, human immunodeficiency virus, Treponema pallidum. 2

3 44 Introduction The potential for transmission of infections during surgical procedures or gastrointestinal (GI) endoscopy (gastroscopy and enteroscopy) is a matter of concern for both physicians and patients (1-7). Currently, transmission of infections due to GI endoscopy is a rare event with a transmission rate of one in 1.8 million (1). Further, documentation of transmission of viral infections by GI endoscopy is difficult due to long incubation periods of these infections and the presence of no or minimal symptoms in these patients (1). Pre-operative routine testing for the presence of infective pathogens may benefit these patients as it can provide information on the pathogen status and help in subsequent counselling, care and access to treatment, where applicable (3, 7). Also, this type of testing can provide an opportunity for healthcare professionals to implement measures to prevent transmission of infections to hospital workers involved in the surgery (1, 7). It has been suggested that pre-operative testing for infective pathogens involves a large number of investigations that are often expensive, rarely detect major abnormalities and may cause unnecessary delays or cancellation of surgeries leading to an increase in medico-legal liability (3). Currently in China, serological methods are used for routine screening of blood samples of patients undergoing surgery and GI endoscopy for blood-transmissible infections including Treponema pallidum (syphilis), hepatitis B virus (HBV), hepatitis C virus (HCV) and human immunodeficiency viruses 1 and 2 (HIV 1/2) (1-3, 7). A screening algorithm defines specific tests and testing procedures to be followed during the screening of infective pathogens in each hospital facility, which helps in maintaining consistency of results and decisions for the management of patients with confirmed positive 3

4 results (8). All proposed algorithms should have a high level of sensitivity and specificity for testing the blood-transmitted infections to ensure operative safety (9-11). The selection of appropriate algorithms and assays is a critical part of the screening program, and enzyme immunoassays (EIA) and chemiluminescent immunoassays (CLIAs) are currently the most frequently employed in these serological screening algorithms (9-13). While most of the currently used screening algorithms are still based on the use of gold immunoassays and colloidal tests (GICT), the use of algorithms based on automated chemiluminescent methods can provide several important advantages (8, 14). Among the commercially available CLIAs, the BIO-FLASH technology (Biokit, Barcelona, Spain) provides automated two-step CLIAs for the qualitative measurement of HBV surface antigen (HBsAg) and antibodies against HCV, HIV 1/2 and T. pallidum (syphilis) in human serum or plasma as an aid in the proper serodiagnosis of HBV, HCV, HIV 1/2 and syphilis (14). The present observational cross-sectional study assessed the advantages of the newly proposed algorithm based on BIO-FLASH technology over the current GICT algorithm using manual colloidal selenium one-step immunoassay strips in terms of sensitivity, specificity and workflow characteristics for the screening of HIV, HBV, HCV and syphilis in patients undergoing GI endoscopy. 83 Materials and methods 84 Study design This observational, cross-sectional study was conducted at the Department of Blood Transfusion, General Hospital of Beijing Military Region, Beijing, China between 14 th 18 th 4

5 April 2015 and 1 st 31 st July The study was performed in accordance with national legislation and the Declaration of Helsinki (revised in 2000). Patients undergoing GI endoscopy were invited to participate in this study. Blood samples of the participating patients were collected by venepuncture. Serum was separated from the collected blood samples and tested for HIV, HCV, HBV and syphilis infection markers. Retrospective samples of patients admitted to the hospital between 8 th January and 16 th March 2015 were analysed for HIV as no HIV-positive patients enrolled in the study. A representative screening algorithm flowchart for detection of infection markers is shown in Figure 1. The main aim of the proposed algorithm for screening of serum samples for infection markers was the inclusion of the two-step BIO-FLASH technology. Initial reactive samples were automatically repeated using BIO-FLASH and the repeated reactive samples were subsequently analysed using GICT on the same day. Any discordant results were analysed using Enzyme-linked immunosorbent assay (ELISA) on Day 2 (Figure 2). In the conventional manual algorithm, the early stages of screening were performed using GICT (as a regular laboratory procedure). Positive results were confirmed using ELISA on Day 2 and 3 (Figure 2) and any discordant results between ELISA and GICT were confirmed using western blot. All tests were performed as per the manufacturer s instructions. Patient s diagnostic results were stratified into five different categories for all markers. Group A included all patients with negative results in the screening test; Group B included all patients with positive results in the screening test but confirmed negative in a second analysis; Group C included all patients with positive results in the screening test and confirmed positive in a second analysis; Group D included all the patients with discordant 5

6 results between the two early tests and confirmed negative result by a third method and Group E included all patients with discordant results between the two early tests and confirmed positive result with a third method (Figure 2). 113 Automated BIO-FLASH technology Automated two-step CLIAs were used for qualitative measurement of IgG and IgM antibodies against HIV-1 and HIV-2 (BIO-FLASH anti-hiv 1+2, Biokit, Barcelona, Spain), IgG antibodies against HCV (BIO-FLASH anti-hcv, Biokit, Barcelona, Spain), IgG and IgM antibodies against T. pallidum (BIO-FLASH Syphilis, Biokit, Barcelona, Spain) and HBsAg (BIO-FLASH HBsAg, Biokit, Barcelona, Spain) in the serum samples using the BIO-FLASH instrument. Time to first automated results was 30 minutes and 64 tests/hour were reported after the first hour. Initial reactive samples were automatically repeated and only repeated reactive samples were considered positive. 122 Manual gold immunoassays and colloidal tests Manual colloidal selenium one-step immunoassay strips were used for qualitative measurements of antibodies against HIV-1 and HIV-2 (Alere Determine HIV-1/2, Alere Medical Co., Ltd, Waltham, USA) and T. pallidum (Alere Determine Syphilis, Alere Medical Co., Ltd, Waltham, USA) in the serum samples. Individual results were visually read after 15 minutes each. Similar strips were also used for qualitative measurements of antibodies against HCV (Ying Ke Xin Chuang [XiaMen] Co., Ltd, Xiamen, China) and detection of HsBAg (Ying Ke Xin Chuang [XiaMen] technology Co., Ltd, Xiamen, China) where individual results were visually read after 10 and 15 minutes each, respectively. 6

7 131 Enzyme-linked immunosorbent assay Confirmatory tests were performed using ELISA. Two different confirmatory tests were used for each infection marker: for HIV (1/2), Ab ELISA (Beijing Wantai Biological Pharmacy Enterprise, Beijing, China) and Genscreen HIV 1/2 v.2 (Biorad, California, USA); for HCV, ELISA (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd., Beijing, China) and Murex anti-hcv 4.0 (Abbott Murex, Kent, UK); EIA kit for the detection of T. pallidum (Shanghai Kehua Bio-engineering Co., Ltd, Shanghai, China) and syphilis EIA (Abbott Murex, Kent, UK); for HBsAg, ELISA (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd, Beijing, China) and Murex HBsAg 3.0 (Abbott Murex, Kent, UK). 140 Western blot Discordant results between BIO-FLASH and GICT and ELISA were confirmed using western blot in all cases except for HBV, which was confirmed using five different HBV infection markers (HBeAg, IgM anti-hbc, total anti-hbc and anti-hbs). For the detection of antibodies against HCV, the HCV antibody detection kit was used (MP Biomedical Asia Pacific Pte, Ltd, Singapore). For the detection of T. pallidum antibodies, the membrane-based test system EUROLINE-WB (Euroimmun Medical Laboratory Diagnostics Stock Company, Beijing, China) and T. pallidum haemagglutination assay (Immutrep Syphilis Test Kits; Zhuhai Xin Mei Trading Co, Ltd, Guangdong Province, China) were used. Antibodies against HIV (1/2) were detected using the MP-WB kit (HIV Blot 2.2 WB, MP Biomedicals Asia Pacific Pte, Ltd, Singapore). 151 Statistical analysis 7

8 Data were analysed separately for each infection marker using the SPSS software version For each test, positive and negative results were defined according to the cut-off values specified in the manufacturer s instructions. The clinical diagnosis of each infection was used as a gold standard which was blinded for laboratory operators. Test sensitivity was defined as TP/(TP+FN), where TP was true positive results and FN was false negative results. Similarly, test specificity was defined as TN/(TN+FP), where TN was true negative results and FP was false positive results. Positive predictive value (PPV) was defined as TP/(TP+FP) and negative predictive value (NPV) was defined as TN/(TN+FN). Sensitivity, specificity and predictive values of both algorithms were evaluated and expressed as percentage (%). The level of agreement between BIO-FLASH and GICT to detect each infection marker at an early stage of algorithms was calculated using kappa coefficient (k-value; 95% confidence interval) without assuming the null hypothesis. The laboratory workflow was evaluated in the early stages for both screening algorithms. The number of tests and handling time for the tasks performed by laboratory personnel per batch of samples (number of samples received per hour) were analysed for both the algorithms. Patient s diagnostic workflow was also analysed for the time (days) and number of visits to the hospital required to obtain the final diagnostic results for the selected infection markers using the two screening algorithms. 169 Results A total of 956 patients aged years (mean age 44.8 ± 18.1 years; 59.8% male) were enrolled in the study. The baseline characteristics of the patients included in the study are shown in Table 1. A total of 176 samples from the hospital frozen repository were used to assess the sensitivity and specificity of both the algorithms for HIV1/2 (Table 1). Sensitivity and specificity of the proposed BIO-FLASH algorithm was 100% for the detection of anti- 8

9 HIV 1/2, whereas for the manual algorithm, sensitivity was 80.0% and the specificity was 98.6% (Table 3). PPV and NPV were 100.0% using the BIO-FLASH (Table 3). Eight discordant results were observed and resolved by confirmatory western blot, of which six positive samples were confirmed negative and two negative samples were confirmed positive (Table 3). The BIO-FLASH showed higher test sensitivity to detect anti-hcv antibodies (100%) than the manual GICT algorithm (75.0%) with 100% specificity for both algorithms. PPV and NPV were 100.0% using the BIO-FLASH (Table 2). One discordant result was obtained using both algorithms, which was confirmed positive using western blot. Similarly, the automated algorithm using BIO-FLASH showed higher sensitivity for the detection of HBsAg (100%) than manual GICT algorithm (94.7%), with 100% specificity for both algorithms (Table 2). PPV and NPV were 100.0% using the BIO-FLASH algorithm (Table 2). One discordant result was obtained using both algorithms which was confirmed as positive, by testing five different HBV infection markers (HBeAg, IgM anti-hbc, total anti- HBc and anti-hbs) (Table 2). For T. pallidum detection, both algorithms were shown to be sensitive and specific (100% in all cases) (Table 2). PPV and NPV were 100.0% using both algorithms (Table 2) and no discordant results were observed. Level of consistency between the two algorithms to detect infection markers was high for all infections: anti-hiv-1/2 antibodies (k-value=1), anti-hcv antibodies (k-value=0.857), T. pallidum (k-value=0.972) and HBsAg (k-value=0.972) (Table 4). Using BIO-FLASH, a total of 909/956 patients were diagnosed negative for all infection markers at the early stages of screening (group A) and 47/956 were diagnosed as positive, of whom 31 participants were confirmed positive (group C) and 16 had discordant diagnosis by GICT. After the 9

10 confirmatory test analysis, 14/16 participants were confirmed negative (group D) and 2/16 participants were confirmed positive (group E) (Figure 2; Table 5). Using the manual GICT screening algorithm, a lower percentage of patients (896/956) were diagnosed negative for all infection markers at the early stages of screening (Group A) and 60 remaining patients were diagnosed as positive of whom 31 were confirmed positive (group C) and 29 were confirmed negative (group B) by confirmatory tests (Figure 2; Table 5). Using the new algorithm with two BIO-FLASH instruments, handling was required only to load the sample batches every hour. The first set of results was generated within 30 minutes of loading the samples and continuous loading of the samples gave the results of the four infectious markers from 30 samples in 60 minutes. The total hands-on time using the new automated BIO-FLASH algorithm was 35 minutes for complete screening analysis including start-up and maintenance compared with the manual GICT which required 240 minutes for generation of the final report (Figure 3). The automated BIO-FLASH screening algorithm also reduced the time required to obtain complete diagnostic results of infection markers and the frequency of hospital visits for patients compared with the regular laboratory procedure using the manual GCIT screening algorithm (Table 5). The BIO-FLASH algorithm required a maximum of 2 days and a maximum of two hospital visits to reach the final diagnosis for all groups compared with the manual GICT screening algorithm which required 6 days for some patients (Figure 2). Complete diagnostic results were obtained for 95.08% of patients as early as 60 minutes after collection of blood samples, 98.32% of patients received the complete diagnostic result on Day 1 (groups A, B and C), and only a 1.67% of patients were required to schedule a second visit (Table 5) with BIO-FLASH. For manual GCIT screening algorithm, 93.72% of patients 10

11 received complete diagnostic results within 60 minutes on Day 1 (group A) while the rest of the patients (6.27%; groups B and C) were required to schedule a second visit to the hospital and received their results on Day 2 (Table 5); no patients were included in groups D and E. 224 Discussion The present study compared the sensitivity, specificity and workflow characteristics of the automated two-step BIO-FLASH technology with the manual GICT for the screening of infection markers in patients undergoing GI endoscopy. The BIO-FLASH screening algorithm showed 100% sensitivity and specificity in the detection of HIV, HCV, HBsAg and T. pallidum markers. The newly proposed BIO-FLASH screening algorithm also reduced the handling time required for the assay, the overall time required to generate the diagnostic results and the number of hospital visits compared with the routinely used manual GICT algorithm for the detection of infection markers in these patients. Automated CLIA analysers are used for routine serological assays in high-volume clinical laboratories. These instruments offer excellent precision and reliability, high-speed throughput, random access and technical simplicity. Although automated CLIAs are gradually replacing the EIAs, there are data from published studies comparing the algorithms using the two techniques (9, 10, 12, 15-17). It has been reported that detection of infection markers requires all screening assays to have a high level of sensitivity and specificity (9-11, 18). The present study demonstrated for the first time that the use of an algorithm based on automated CLIAs before GI endoscopy increased the sensitivity of detection of HIV, HCV and HBV, increased the specificity of HIV diagnosis and maintained the specificity for HCV, HBV and syphilis detection, while allowing a substantial reduction in the hands-on time and the total time required to obtain the diagnostic results. These results are in line with previous 11

12 studies where the use of CLIAs was reported to be highly specific and sensible for the diagnosis of HCV and HBV compared with conventional EIAs (9, 10, 12). It was also reported in the present study that BIO-FLASH showed an increased specificity for the detection of HIV due to a lower number of observed false positive results. False positive results in HIV samples are problematic and it is important to use approaches that minimize the number of biological false positive screening test results. Since biological false positives occur for a variety of reasons, confirmatory tests are necessary during screening. It is generally recommended by the assay manufacturers and health authorities that the serological screening should be repeated twice for all positive samples using the same assay before proceeding to confirmatory tests for these repeatedly reactive samples (11, 19-21). BIO-FLASH assays have been shown to be useful for detection of infection markers although more research is needed about their implementation, acceptability and costs in routine clinical practice (22). An important advantage of the proposed algorithm with automated BIO-FLASH technology reported in the present study was the reduction in the working time of the laboratory technicians. The results of the study indicate that the use of automated BIO-FLASH screening algorithm reduced the operation time of the laboratory technicians by 85% (205 minutes) per day for the detection of infection markers in patients before GI endoscopy, saving up to 24 hours/ week. Further, the use of the proposed algorithms also improved the laboratory workflow in terms of connection with laboratory information system (LIS), availability of historic results, interpretation of results and automation. Importantly, the proposed algorithm also increased the percentage of patients receiving their final report and confirming their GI surgery on the same day of the blood sample collection, avoiding a 12

13 second visit to the hospital. Thus, the use of this algorithm can help in the early detection and management of an infectious disease in case of positive results. Finally, due to increased sensitivity and specificity, the use of the proposed BIO-FLASH screening algorithm reduced the number of samples that needed to be confirmed in an external laboratory. Of note, in this study, no patients were included in groups D and E of the manual GICT algorithm, for whom a confirmatory western blot analysis would be needed. Data documented in our laboratory showed that a total of 82 samples (n = 12 for HCV, n = 29 for syphilis and n = 41 for HIV) in 2015 and 76 samples in 2014 (n = 13 for HCV, n = 7 for syphilis and n = 56 for HIV) were sent to an external laboratory for confirmatory western blot assay accounting for 0.38% and 0.30% of all specimens, respectively. Thus, the use of the currently proposed automated BIO-FLASH technology may be beneficial in the processing of these samples. In conclusion, the present study demonstrated for the first time the advantages of a screening algorithm based on the automated BIO-FLASH technology for the detection of HIV, HCV, HBV and syphilis in patients undergoing GI endoscopy

14 283 Acknowledgements Authors wish to thank Núria Piqué for her editorial assistance and Nishad Parkar, PhD, of Springer Healthcare Communications for providing assistance in English editing of the manuscript. 287 Funding sources 288 This study was funded by Biokit, SA (Barcelona, Spain). 289 Conflict of interest statement 290 The authors declare that there is no conflict of interest

15 292 References Banerjee S, Shen B, Nelson DB, Lichtenstein DR, Baron TH, Anderson MA, Dominitz JA, Gan SI, Harrison ME, Ikenberry SO, Jagannath SB, Fanelli RD, Lee K, van Guilder T, Stewart LE Infection control during GI endoscopy. Gastrointest Endosc 67: Weber P, Eberle J, Bogner JR, Schrimpf F, Jansson V, Huber-Wagner S Is there a benefit to a routine preoperative screening of infectivity for HIV, hepatitis B and C virus before elective orthopaedic operations? Infection 41: Ahmed R, Bhattacharya S Universal screening versus universal precautions in the context of preoperative screening for HIV, HBV, HCV in India. Indian J Med Microbiol 31: Wu H, Shen B Health care-associated transmission of hepatitis B and C viruses in endoscopy units. Clin Liver Dis 14:61-68; viii. 5. Lisgaris MV The Occurrence and Prevention of Infections Associated with Gastrointestinal Endoscopy. Curr Infect Dis Rep 5: Nelson DB, Muscarella LF Current issues in endoscope reprocessing and infection control during gastrointestinal endoscopy. World J Gastroenterol 12: Odigie J, Siminialayi I Perception and attitude of theatre staff to preoperative HIV testing at the University of Port Harcourt Teaching Hospital. Asian Pacific Journal of Tropical Medicine 3:

16 World Health Organisation Screening Donated Blood for Transfusion- Transmissible Infections: Recommendations. Geneva: World Health Organization; , Screening assays. 9. Dufour DR, Talastas M, Fernandez MD, Harris B Chemiluminescence assay improves specificity of hepatitis C antibody detection. Clin Chem 49: Kim S, Kim JH, Yoon S, Park YH, Kim HS Clinical performance evaluation of four automated chemiluminescence immunoassays for hepatitis C virus antibody detection. J Clin Microbiol 46: Sommese L, Iannone C, Cacciatore F, De Iorio G, Napoli C Comparison between screening and confirmatory serological assays in blood donors in a region of South Italy. Journal of clinical laboratory analysis 28: Ismail N, Fish GE, Smith MB Laboratory evaluation of a fully automated chemiluminescence immunoassay for rapid detection of HBsAg, antibodies to HBsAg, and antibodies to hepatitis C virus. J Clin Microbiol 42: Sommese L, Paolillo R, Sabia C, Costa D, De Pascale MR, Iannone C, Esposito A, Schiano C, Napoli C Syphilis detection: evaluation of serological screening and pilot reverse confirmatory assay algorithm in blood donors. Int J STD AIDS 27: Costa M, Canela B, Martinez A, Faraudo S, Garcia M, Lopez M, Hervas C, Rodriguez C, Tous J, Puig J, L T. Evaluation of BIO-FLASH HBV/HCV/HIV 1+2 assays on biokit s BIO-FLASH analyser. 23 rd European Society of Clinical Microbiology and Infectious Diseases Congress 2013; Berlin, Germany. 16

17 Jonas G, Pelzer C, Beckert C, Hausmann M, Kapprell HP Performance characteristics of the ARCHITECT anti-hcv assay. J Clin Virol 34: Myrmel H, Navaratnam V, Åsjø B Detection of antibodies to hepatitis C virus: False-negative results in an automated chemiluminescent microparticle immunoassay (ARCHITECT Anti-HCV) compared to a microparticle enzyme immunoassay (AxSYM HCV Version 3.0). Journal of Clinical Virology 34: Watterson JM, Stallcup P, Escamilla D, Chernay P, Reyes A, Trevino SC Evaluation of the Ortho-Clinical Diagnostics Vitros ECi Anti-HCV test: comparison with three other methods. J Clin Lab Anal 21: Kleinman SH, Lelie N, Busch MP Infectivity of human immunodeficiency virus-1, hepatitis C virus, and hepatitis B virus and risk of transmission by transfusion. Transfusion 49: Alter MJ, Kuhnert WL, Finelli L Guidelines for laboratory testing and result reporting of antibody to hepatitis C virus. Centers for Disease Control and Prevention. MMWR Recomm Rep 52:1-13, 15; quiz CE Kiely P, Walker K, Parker S, Cheng A Analysis of sample-to-cutoff ratios on chemiluminescent immunoassays used for blood donor screening highlights the need for serologic confirmatory testing. Transfusion 50: Dow BC Microbiology confirmatory tests for blood donors. Blood Rev 13: McGowan DR, Norris JM, Smith MD, Lad M Routine testing for HIV in patients undergoing elective surgery. Lancet 380:e

18 356 Tables Table 1. Baseline characteristics of the total patient population included in the study and the HIV + patient population. Overall HIV-1/2 + HCV + T. pallidum + HBsAg + No. of patients, n (%) 956 (100) 0 (0) 4 (0.41) 14 (14.6) 20 (2.09) Age, years Range Mean ± SD 44.8 (18.1) Sex, n (%) Male 384 (40.2) 0 (0) 1 (25.0) 5 (35.7) 9 (45.0) Female 572 (59.8) 0 (0) 3 (75.0) 9 (64.3) 11 (55.0) HIV + patient population No. of patients, n (%) 176 (100) 24 (13.6) Age, years Mean ± SD 45.5 (12.5) Sex, n (%) Male - 11 (45.8) Female - 13 (54.2) HBsAg, hepatitis B virus surface antigen; HCV, hepatitis C virus; HIV, human immunodeficiency virus 18

19 Table 2. Comparison of BIO-FLASH and GICT diagnostic algorithms for HIV, HCV, syphilis and HBV marker detection. 362 Clinical gold Algorithms standard diagnosis Sensitivity Specificity PPV NPV Positive Negative (%) (%) (%) (%) HIV 1/2 BIO-FLASH -screening algorithm BIO-FLASH anti-hiv Positive 0 0 Negative GICT-screening algorithm Alere Determine HIV HCV Positive 0 0 Negative BIO-FLASH -screening algorithm BIO-FLASH anti-hcv Positive 4 0 Negative GICT-screening algorithm HCV (XiaMen) Positive 3 0 Negative Syphilis (T. pallidum) BIO-FLASH -screening algorithm BIO-FLASH Syphilis Positive 12 0 Negative GICT-screening algorithm Alere Determine Syphilis HBsAg Positive 12 0 Negative BIO-FLASH -screening algorithm BIO-FLASH HBsAg Positive 20 0 Negative GICT-screening algorithm HsBAg (XiaMen) Positive 18 0 Negative HBsAg, hepatitis B virus surface antigen; HCV, hepatitis C virus; HIV, human immunodeficiency virus NPV, negative predictive values; PPV, positive predictive values 19

20 Table 3. Comparison of results with BIO-FLASH and GICT screening algorithms with clinical diagnosis (gold standard) for HIV detection using an additional panel of samples (n = 176) not included in the study. 366 Clinical Diagnosis Sensitivity Specificity PPV NPV HIV (Gold Standard) (%) (%) Positive Negative (%) (%) BIO-FLASH -screening algorithm BIO-FLASH anti-hiv 1+2 Positive Negative GICT-screening algorithm Alere Determine HIV-1/2 Positive Negative HIV, human immunodeficiency virus; PPV: positive predictive values; NPV, negative predictive values 20

21 Table 4. Level of consistency (k-value) between BIO-FLASH and GICT screening algorithms for the diagnosis of infectious markers (HIV, HCV, Syphilis and HBV). 369 GICT-screening algorithm HIV BIO-FLASH -screening algorithm Kappa-value Positive Negative Total Positive Negative Total HCV Positive Negative Total Syphilis (T. pallidum) Positive Negative Total HBsAg Positive Negative Total GICT, gold immune assays and colloidal tests; HBsAg, hepatitis B virus surface antigen; HCV, hepatitis C virus; HIV, human immunodeficiency virus 21

22 Table 5. Comparison of diagnostic characteristics between CLIAs BIO-FLASH and GICT screening algorithms. Patient groups BIO-FLASH -screening algorithm Time to Day of No. of % of results final hospital pts (hours) report visits % of pts GICT-screening algorithm Time to Day of results final (hours) report No. of hospital visits A B C D E Pts, patients 22

23 373 Figure Legends Figure 1. Representative screening algorithm flowchart for detection of infection markers at the General Hospital of Beijing Military Region, Beijing (China). Positive and negative samples refer to any evaluated markers. Figure 2. Comparison of screening algorithms and tests results for automated chemiluminescent BIO-FLASH screening algorithm (new proposed laboratory procedure) and manual gold immune assays and colloidal tests (GCIT) screening algorithm (regular laboratory procedure). Positive and negative samples refer to any evaluated markers. Figure 3. Hands-on time per day using automated chemiluminescent BIO-FLASH screening algorithm (new proposed laboratory procedure) and manual gold immune assays and colloidal tests (GCIT) screening algorithm (regular laboratory procedure)

24 385 Figures 386 Figure

25 389 Figure

26 392 Figure